private LineBreakClass MapClass(CodePoint cp, LineBreakClass c)
{
// LB 1
// ==========================================
// Resolved Original General_Category
// ==========================================
// AL AI, SG, XX Any
// CM SA Only Mn or Mc
// AL SA Any except Mn and Mc
// NS CJ Any
switch (c)
{
case LineBreakClass.AI:
case LineBreakClass.SG:
case LineBreakClass.XX:
return(LineBreakClass.AL);
case LineBreakClass.SA:
UnicodeCategory category = CodePoint.GetGeneralCategory(cp);
return((category is UnicodeCategory.NonSpacingMark or UnicodeCategory.SpacingCombiningMark)
? LineBreakClass.CM
: LineBreakClass.AL);
case LineBreakClass.CJ:
return(LineBreakClass.NS);
default:
return(c);
}
}
public void MoveNext()
{
this.CharsConsumed += this.charsConsumed;
var codePoint = CodePoint.DecodeFromUtf16At(this.source, this.CharsConsumed, out this.charsConsumed);
this.CurrentType = CodePoint.GetGraphemeClusterClass(codePoint);
}
private static JoiningType GetJoiningType(CodePoint codePoint, uint value, UnicodeCategory category)
{
var type = (JoiningType)(value & 0xFF);
// All others not explicitly listed have joining type U
if (type == JoiningType.NonJoining)
{
// 200C; ZERO WIDTH NON-JOINER; U; No_Joining_Group
// 200D; ZERO WIDTH JOINER; C; No_Joining_Group
// 202F; NARROW NO-BREAK SPACE; U; No_Joining_Group
// 2066; LEFT-TO-RIGHT ISOLATE; U; No_Joining_Group
// 2067; RIGHT-TO-LEFT ISOLATE; U; No_Joining_Group
// 2068; FIRST STRONG ISOLATE; U; No_Joining_Group
// 2069; POP DIRECTIONAL ISOLATE; U; No_Joining_Group
if (codePoint.Value is 0x200C
or 0x200D
or 0x202F
or 0x2066
or 0x2067
or 0x2068
or 0x2069)
{
return(type);
}
// Those that are not explicitly listed and that are of General Category Mn, Me, or Cf have joining type T.
if (category is UnicodeCategory.NonSpacingMark or UnicodeCategory.EnclosingMark or UnicodeCategory.Format)
{
type = JoiningType.Transparent;
}
}
return(type);
}
/// <summary>
/// Initializes a new instance of the <see cref="JoiningClass"/> struct.
/// </summary>
/// <param name="codePoint">The codepoint.</param>
public JoiningClass(CodePoint codePoint)
{
UnicodeCategory category = CodePoint.GetGeneralCategory(codePoint);
uint value = UnicodeData.GetJoiningClass(codePoint.Value);
this.JoiningType = GetJoiningType(codePoint, value, category);
this.JoiningGroup = (JoiningGroup)((value >> 16) & 0xFF);
}
public LineBreakEnumerator(ReadOnlySpan <char> source)
: this()
{
this.source = source;
this.pointsLength = CodePoint.GetCodePointCount(source);
this.charPosition = 0;
this.position = 0;
this.lastPosition = 0;
this.currentClass = LineBreakClass.XX;
this.nextClass = LineBreakClass.XX;
this.first = true;
this.lb8a = false;
this.lb21a = false;
this.lb22ex = false;
this.lb24ex = false;
this.lb25ex = false;
this.alphaNumericCount = 0;
this.lb31 = false;
this.lb30 = false;
this.lb30a = 0;
}
/// <summary> /// Initializes a new instance of the <see cref="BidiClass"/> struct. /// </summary> /// <param name="codePoint">The codepoint.</param> public BidiClass(CodePoint codePoint) => this.bidiValue = UnicodeData.GetBidiData(codePoint.Value);
/// <summary>
/// Advances the enumerator to the next element of the collection.
/// </summary>
/// <returns>
/// <see langword="true"/> if the enumerator was successfully advanced to the next element;
/// <see langword="false"/> if the enumerator has passed the end of the collection.
/// </returns>
public bool MoveNext()
{
this.Current = CodePoint.DecodeFromUtf16At(this.source, 0, out int consumed);
this.source = this.source.Slice(consumed);
return(consumed > 0);
}
